Andy, as a fellow vlogger (and somewhat of a newbie vlogger) I always appreciate the visual clarity of your videos. Could you share with us what equipment and software you use and especially what your final compression settings are… how do you get such clean looking videos yet still keep the file sizes manageable?
Any suggestions might be appreciated by more folks than just myself.
Good questions, Bill. Let’s start with the equipment, because that’s pretty straightforward.
I have two digital cameras I use for video blogging. Most of my videos are shot with a five megapixel Canon Powershot A610. It’s a medium size camera designed for taking still digital photos, but also allows you to shoot video. Its video mode will capture footage in two sizes: 640×480 and 320×240. You can also select between two frame rates: 15 frames a second and 30 frames a second. The more frames you have, the smoother the image quality. But more frames also means bigger video files – I’ll talk about this more in the compression section below.
My other digital camera is an eight megapixel Konica-Minolta Dimage A-200. It looks and feels like a 35mm SLR camera rather than a compact, so it’s good for people used to having a sizable camera. It also lets you shoot in two different sizes and frame rates. Visually, I find the image quality better on this camera, and it lets you adjust the manual zoom while shooting, so it feels more like a real video camera. But it has two major drawbacks: its autofocus is unreliable in video mode, often going out focus. Even worse, the autofocus motor is so loud, you can hear its gears whizzing in your footage, particularly when recording in quiet situations. So generally I use the Dimage for still photography and the Canon for video.
Regarding software, my weapon of choice is Final Cut Pro. I use version 4.5 HD, and have been a fan of FCP since I first edited with it six years ago when version 2.0 came out. Final Cut is amazing software. You could literally edit a Hollywood movie on it – and directors like Robert Rodriguez and Steven Soderbergh have done just that. Having said that, it can also be rather daunting for first-time vloggers, which is why many people use software like iMovie instead.
Another reason I like Final Cut is because it allows me to be very specific as to how I want to compress a video. Compression is simply a technique for saving the movie file so that the file size gets reduced. This is accomplished through very complex algorithms which can probably be explained a heck of a lot better by someone more technical than I. Software like iMovie will often have a limited number of compression modes – if I recall correctly – while Final Cut Pro lets you tweak your compression incrementally and experiment with different results. Similarly, you could purchase Quicktime Pro to do the same thing.
As an example, I’ve uploaded a video clip compressed in a variety of ways, so all of you can take a look and them and compare the results. I started with an uncompressed 10 megabyte file shot on my Canon A610, which shoots video in .avi format. It was recorded at 30 frames per second in full-size mode (640×480 pixels).
10 megabyte avi file, no compression, 640×480 pixels, 30 FPS.
Because the video is uncompressed, the quality is pretty good – it’s very large, yet not clunky or muddy looking. But high quality comes at a cost. Even though the clip is only 5.5 seconds long, it’s a 10 megabyte file, which translates to around two megabytes per second. If you posted an uncompressed avi video that’s five minutes long, you’re then talking about a 600 megabyte file – not exactly the kind of thing most people would be able to download. That’s why you need to compress most videos before posting them online – otherwise, they’re just too big to deal with.
To begin, I start by opening the file in Final Cut Pro, then finding “export” under the File dropdown menu. Here, there are two options. I can just convert it to Quicktime without specifying any type of compression, or I can select “using Quicktime conversion,” which allows you to compress it in a variety of ways. Converting from avi to Quicktime in itself doesn’t really change the file size, though. Take a look at this version of the video It’s saved in Quicktime, 320×160 pixels in size, but without any additional compression. Even though the video is much smaller in size, this reduces the file by only a fraction – down to 9.6 megabytes, a four percent reduction.
Since this didn’t change the file very much, I went ahead and compressed the video using several different techniques:
Reducing frames to 15 frames per second. Here I’ve exported the video using Quicktime conversion. Before saving it, Final Cut Pro lets me select a “options” button, which opens up yet another menu that lets tweak the compression settings. One of these settings is the frame rate. Normally the video plays at 30 frames a second. By changing it to 15 frames a second, it reduces the number of images saved in the video file, in this case cutting the file size nearly in half to 5.4 megabytes. The video doesn’t play as smoothly as 30 frames a second, but still looks tolerably good. In contrast, here’s a version of the same video at 10 frames per second. Notice how the smoothness of the video drops to an intolerable level. Even though the video is now compressed to 3.2 megabytes, the reduced frame rate really hurts the viewing experience. It almost feels like watching a film from the silent era – and not in a positive sense. So, try not to go less than 15 frames per second.
Using 3ivx compression at 15 frames per second. Whether you use FCP or Quicktime Pro, the software comes with a variety of compression algorithms, or CODECs, you can select from. You can also buy CODECs that are optimized for online downloads. I use one called 3ivx. It’s easy to buy online and install, and it compresses video quite nicely. Compare the 3ivx version with the previous version, both of which are set at 15 frames per second. While the basic version is 5.4 megabytes, the 3ivx version is only 1.8 megabytes. That’s more than 80% smaller than the original file.
3ivx codec, changing video size to 160×120 pixels, 15 frames per second. You can also change the size of your video. When exporting it, I clicked the option button and selected the size option, then reduced the height and width by half. This makes a much smaller video window and reduces the file size even further to 1.3 megabytes. Cutting the video screen size, of course, affects the viewing experience, as you can see by comparing these two screen sizes:
Two screen sizes: 320×240 pixels versus 160×120 pixels
3ivx codec, 15 fps, 8 bit mono. Don’t forget you can also compress a video’s audio to reduce file size. Once again, you can reach this menu by exporting as quicktime conversion and clicking the options button; you’ll see an audio option along with the others. The default setting for audio is 48khz, 16 bits and stereo. These settings may not mean much to you, but essentially, it allows for high-quality audio. This setting makes sense when you are using professionally recorded sound or music in your video, but if you’re just using the mic on a digital camera, chances are it’s overkill. For example, my camera records audio only at 8 bits and mono, so not lowering the settings is a waste of bandwidth. By switching the settings to 8 bits, 16khz and mono, the file size drops to 980 bytes – less than one megabyte.
3ivx codec, 15 fps, 8 bit mono, 160×120 pixels. Last but not least, we have a video clip that’s got all of these compression tricks used at once. It’s using the 3ivx CODEC, at 15 frames per second, the audio quality down to 8 bit mono, and the screen size cut down to 160×120. The result is a video file that’s an amazingly small 292 kilobytes – a whopping 97% smaller file size than the original. And amazingly, the final quality isn’t so terrible. In fact, for people who aren’t used to larger screen sizes, it’s downright tolerable. Compression techniques like this allow you to upload and download video clips through a dialup connection – and this can come in very handing when working in parts of the world that have slow Internet access.
For those of you who compares all of these techniques more easily, here’s a handy table I created. Hope this helps… -andy